Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 53, 1921
This text is also available in PDF
(3 MB) Opens in new window
– 386 –

Art. XLIV.—Further Studies on the Prothallus, Embryo, and Young Sporophyte of Tmesipteris.

Read before the Philosophical Institute of Canterbury, 1st December, 1920; received by Editor, 31st December, 1920; issued separately, 12th August, 1921.]

Plate LXIII.

The first accounts of the prothallus of Tmesipteris to be published were those contained in the two papers of Professor A. A. Lawson (11, 12), in the latter of which the author also described the prothallus of Psilotum. In the same year Darnell-Smith (3) published an account of the prothallus of Psilotum, and described his successful attempts to germinate the spores experimentally. Lawson's two accounts relate to the mature prothallus and sexual organs of both Tmesipteris and Psilotum, there being shown to be a more or less close resemblance between the two plants with regard to the gametophyte generation. A single embryo of Tmesipteris was figured also in his first paper.

In the following year I published an account (7) of the prothallus and young plant of Tmesipteris, based on abundant material obtained in the wet forests of Westland, New Zealand. The development of the sexual organs and of the embryo was described, but in the case of the latter the series obtained was incomplete, although it indicated the absence of root, suspensor, and cotyledon. The main object of the present paper is to trace more fully the development of the embryo and of the young plant. The absence of a root organ from the adult plant, and its probable absence also from the embryo, together with the discovery of the rootless Rhyie plant fossils in the Scottish Early Devonian (8, 9, 10), gives to the Tmesipteris embryo and young plant an exceptional interest. Although the series of embryos studied for the purpose of the present paper is still not altogether complete, the results obtained seem to be such as to warrant immediate publication.

The new material on which this account is based has, as before, all been obtained in Westland. It embraces about seven hundred prothalli in all, many of which bore embryos of different ages and attached plantlets, and also a full series of detached plantlets.

It may be mentioned that the embedding in paraffin was practically all done at home by the aid of the simple brass table illustrated by Chamberlain on page 14 of the third edition of his Methods in Plant Histology (Chicago, 1915), the results obtained being quite satisfactory. The stain used throughout was Delafield's haematoxylin, this being chosen on account of its clear differentiation of embryos. In some cases safranin was used in conjunction with it. The drawings were made with the aid of a Leitz camera lucida. I have to thank Dr. Charles Chilton, of the Canterbury College Biological Laboratory, for permission to work in the laboratory from time to time, and for his interest in my work.

– 387 –


The climate of Westland is a continuously wet one, there being practically no really dry periods at any season of the year. For example, at Hokitika, on the coast, as shown in the Meteorological Office records, for the ten years 1909 to 1918 the average annual rainfall was 117.36 in., the lowest for any one year being 100.97 in. and the highest 134.32 in. A detailed examination of these records shows that during this decade twice only was the monthly rainfall less than 2 in., while generally speaking the annual total was fairly evenly distributed over the twelve months, and no one month in the year showed usually a markedly less rainfall than any other. As a general rule, also, the Westland climate is characterized by the absence of dry winds. On the main mountain-ranges, which run more or less parallel with the coast, the rainfall is, of course, much heavier than at sea-level, this being especially so in the gorges and on the lower flanks of the ranges. For example, at Otira, at the western end of the tunnel which pierces the main divide, on the Midland Railway, lying at an altitude of 1,255 ft., the average rainfall for the five years 1914 to 1918 was 198.73 in.

On account of the wet climate and constantly high humidity the whole district from sea-coast to the mountain-ranges is covered with heavy forest, and the growth of ferns and other cryptogamic plants is luxuriant both epiphytically and on the floor. Tree-ferns, especially Dicksonia squarrosa, are abundant in the lowland forest and up to the bases of the mountains. Away from the coast Metrosideros lucida (the southern rata) is a common member of the forest, and its large much-branched and irregularly-growing trunks frequently show thick accumulations of epiphytic humus with colonies of Pteridophytes.

In the coastal forest Tmesipteris occurs abundantly on the stems of Dicksonia, being frequently accompanied in this station by Lycopodium Billardieri var. gracile and by the two filmy ferns Hymenophyllum ferrugineum and Trichomanes venosum. In this part of the district the mature plants of Tmesipteris also occur, but far less frequently, in the forks of large trees or even on the ground. Colonies of young plants can often be found on the tree-fern stems, the youngest plants being invariably at the uppermost limit of the colony. The most favourable place for the germination of the spores is clearly that part of the stem in which the bases of the frondstipites are beginning to form a firm but not too dense substratum with an accumulating humus by the extension upwards of the tree-fern's clothing of aerial rootlets. As the Dicksonia grows in height the Tmesipteris plantlets extend upwards, those farther down the stem exhibiting progressively older stages of development. These are favourable places for finding prothalli, in some cases in relative abundance. I have dissected out from selected portions of tree-fern stems a total of considerably over one hundred prothalli. In this coastal forest, and even farther inland wherever Dicksonia occurs commonly, one can always be sure of obtaining the young plants and prothalli, although the work of dissecting them out from the tree-fern-stem surface is generally tedious and requires considerable care. Undoubtedly the easiest places in which to find the prothalli of Tmesipteris are the large overhanging trunks of the rata, where, as has been mentioned previously, humus frequently accumulates to a considerable depth on the upper sides of the trunks and lower limbs and in the crevices, within easy reach of the ground. Here colonies of large mature plants are to be met with, the humus being permeated with their rhizomes. The young plants, if present, occur quite indiscriminately, there being an absence of the useful grading which is

– 388 –

invariably to be found on tree-fern stems, but with the different advantage that the humus lacks the irritating entanglement of tough aerial tree-fern rootlets and is easily crumbled down. I have found that Tmesipteris occurs in this station more particularly at middle altitudes in the district, on the lower parts of the mountain-flanks and in the valleys, where the rainfall is even heavier than at the sea-coast; and, although I have not often found large colonies of the young plants in these situations, I feel sure that a systematic investigation of these large overhanging tree-trunks would show that the plantlets occur not infrequently. From one particular rata in the valley of the Greenstone River, on the lower parts of the Hohonu Range, at an altitude of about 1,300 ft., I took home on each of three occasions a parcel of humus and secured altogether no less than 580 prothalli.

On the flanks of the ranges Tmesipteris is frequently to be met with growing in the thick humus on the forest-floor either as single plants or in colonies, although here, owing to the dense accumulation of undecayed vegetable debris, germination of the spores probably does not take place frequently.

The Prothallus.

I have already given (7) a fairly complete account of the form and structure of the prothallus and development of the sexual organs. As a result, however, of the study of the very large number of prothalli since discovered by me there are some additional facts to be noted.

Of the total number of prothalli found on the one rata, as mentioned in the previous section of this paper, about one hundred were quite young—that is, they were from 1 mm. to 2 mm. in length. A good many of these showed the original spore still attached. The youngest found was just under 1 mm. in length, and is shown in general view in fig. 1. It had developed no sexual organs. Its lower two-thirds was brown in colour, as also were the rhizoids, but the head was colourless. Throughout the brown region the cell-walls showed very distinctly, as is always the case in the Tmesipteris prothallus, while the cells themselves each contained a circular compact fungal skein. At the uppermost limit of the brown region, however, the fungal skeins were thin and delicate. The browning of the cell-walls extended slightly beyond the point reached by the fungus, and also a slight general tinge was beginning to show on the left side of the head. This brown or almost golden-yellow colour is characteristic of both young and old prothalli of Tmesipteris, except for the actual head, and it seems to arise very early in the development. Darnell-Smith (3, p. 86) notes that in Psilotum the first cells formed on germination of the spore are light brown in colour, and that the fungus begins to infect the prothallus at the three-cell stage. He also notes (ibid., p. 88) that the small bulbils borne on the older prothalli are colourless in the early stages of growth, but later become brown. The cells of the head of the prothallus shown in fig. 1 were large and bulging, and contained much starch. There was no basal filament of a single cell in thickness, as is sometimes to be seen even in older prothalli (7, figs. 1, 11), but the prothallus began immediately from the spore to increase gradually in width by cell-multiplication. The fungus was present in the lowest cells of the prothallus and in the spore also. The basal region of another young prothallus with the spore attached is shown in fig. 2. The wall of the spore is thick, and its outer surface pimpled.

Older prothalli not infrequently show the basal end intact, and in some cases the spore can still be seen attached. Fig. 4 shows in general view

– 389 –
Picture icon

Fig. 1.—A very young complete prothallus in general view. × 110.
Fig. 2.—Basal end of a young prothallus in general view, showing the spore. × 270.
Fig. 3.—Basal end of the prothallus shown in fig. 4. × 110.
Fig. 4.—Medium-aged prothallus in general view, showing twofold forking of the head, and also the distribution of the fungus. × 35.

– 390 –

a medium-aged prothallus which is beginning to fork, and fig. 3 its basal end. Fig. 5 is a median longitudinal section of the lower regions of a still older prothallus, in which the gradual extension in width in a series of gentle swellings is clearly seen. The latter prothallus had preserved the unbranched carrot form unusually long (fig. 6), and was rather more attenuated in form than usual, but it illustrates well the manner in which the Tmesipteris prothallus always increases in girth. That shown in fig. 4 is more typical. In the latter the first forking of the head had already taken place, and the two large swollen apices were in the act of forking again. The shaded area in the main body shows the extent of the fungal distribution, the two apices being quite clear.

The prothalli vary a good deal in both length and thickness. Measurements of mature specimens showed variations in thickness from 0.3 mm. to 1.25 mm., and an extreme length of 18 mm. has been observed Generally speaking, it is the more attenuated prothalli which show the greater tendency to an extension in size by a second forking of the apices, the thicker individuals seldom appearing to fork more than once unless one of the first branches has ceased to grow. The stouter prothalli frequently possess much-swollen heads. The apex of the attenuated form was given in longitudinal section in my previous paper at fig. 21, while that of the stouter form is shown in the present paper in longitudinal section in fig. 7, and in transverse section in fig. 8. There is always a single apical cell of the same form as is found throughout the life of the sporophyte. I have found that serial sections of large thick prothalli may show the presence of several fertilized archegonia and very young developing embryos in close proximity to one another, but I have seldom found more than one plantlet attached to a single prothallus.

Picture icon

Fig. 5.—Longitudinal section of the basal end of the prothallus shown in fig. 6. × 63.
Fig. 6.—General longitudinal section of a prothallus, showing its gradual extension in width, and also an attached plantlet. × 12.

The close similarity in general appearance between the prothallus and the very young sporophyte must be noted. This, of course, arises mainly from the fact that the prothallus possesses an extended and branched chlorophyll-less body with a fairly regular radial growth, and that the young sporophyte consists at first of a simple branching rhizome devoid of appendages, both being brown in colour and covered with the same long brown rhizoids. It is sometimes quite impossible to be sure to which of the two a fragment belongs until it is closely examined under the microscope. This similarity is more marked in the case of Tmesipteris than in any other Pteridophyte The young adventitiously-produced plantlet of Psilotum is also similar in appearance to the prothallus.

– 391 –

Some additional figures illustrating the development of the sexual organs are given in the present paper. That of the antheridium is easily followed (figs. 9 to 19). It begins at an early stage to project from the surface, and at the mature stage does so very strongly. The number of sperm-cells is much less than in either Lycopodium or in the Ophioglossaceae.

Picture icon

Fig. 7.—Longitudinal section of a stout apex of a prothallus, showing the apical cell. × 100.
Fig. 8.—Transverse section of a stout apex of a prothallus, showing the apical cell. × 100.
Fig. 9–17.—Developmental series of antheridia in longitudinal section. × 100.

Spermatogenesis was not followed, but Lawson states that the spermatozoids of Psilotum are multiciliate (12, p. 105). Two additional figures illustrating the development of the archegonium are also given (figs. 20, 21). It is clear that there is here no basal cell. The demonstration of neck-canal cells and the ventral-canal cell which was left over from my previous

– 392 –

paper I have not been able to determine satisfactorily. As with the antheridium, the archegonium projects strongly from the surface, practically only the venter being sunk.

Picture icon

Fig. 18.—Transverse section of a large antheridium, showing the main divisions. × 250.
Fig. 19.—The same antheridium as in fig. 18, showing the opercular cell. × 250.
Fig. 20—Longitudinal section of a stout apex of a prothallus, showing two young archegonia but not the apical cell. × 75.
Fig. 21.—Longitudinal section of a medium-aged archegonium. × 170.

Adventitious Prothallial Buds.

Three instances of portions of old prothalli bearing small adventitious buds were noticed. On one of these three young buds had been formed, these being shown in longitudinal section in figs. 22 to 24. Fig. 24 shows the actual apex of the bud marked X in fig. 23, from which it is clear that there is a single apical cell. On another fragment of an old prothallus three buds in different stages of development were found, one (fig. 25) being quite young, and the other two (figs. 26, 27) much older. The two latter became detached from the prothallus. Sexual organs were present on both these fragments, so that their prothallial nature is beyond question. The buds were in every case packed with starch, and fungal coils were present in the prothallial cells which immediately adjoined them. The buds arise from the superficial cell-layer of the prothallus, but it is not quite clear whether one or two of these cells are concerned in their formation. The sections shown in figs. 22 and 23 make it appear that the

– 393 –

buds arise from two cells, but the bud shown in general view in fig. 26, judging from the old point of attachment at its base, seems to have arisen from a single cell The two largest buds found (figs. 26, 27) were somewhat brown in colour in their lower portions, the cell-walls being strongly marked as in the case of prothalli formed from the spores. The cells in this region showed the presence of fungal coils. From fig. 22 it is clear that the fungus had infected this particular bud through a rhizoid Nothing was found to illustrate the further history of these structures.

Picture icon

Fig. 22, 23.—Young prothallial adventitious buds in longitudinal section. × 75.
Fig. 24.—Longitudinal section of the bud marked × in fig. 23, showing its apical cell. × 75.
Figs. 25–27.—Prothallial buds of different ages in general view. Fig. 25 × 75; fig. 26 × 44; fig. 27 × 30.

Darnell-Smith notes (3, p. 87) that the prothallus of Psilotum (presumably P. triquetrum) frequently bears small bulbils which are carried upon short stalks one cell in width. I have found on old rhizomes of P. triquetrum collected on the scoria islet of Rangitoto, Auckland Harbour, an abundance of the bulbils which were first described by Solms Laubach (14). But no corresponding structures have been observed on the rhizomes of Tmesipteris, so that it is interesting to find that the prothallus of the latter bears under certain conditions small superficial buds. The occurrence of adventitious buds in both generations of the Psilotaceae increases the similarity between them noted above.

– 394 –

The Embryo.

A. General Observations.

From the study of the limited number of embryos described in my former paper I drew the following conclusions. The first wall to be formed in the zygote, the basal wall, is transverse to the direction of the axis of the archegonium, and separates the embryo into its two main regions, the hypobasal (lower) region wholly giving rise to the foot, and the epibasal wholly to the shoot. There is no suspensor, cotyledon, or root. The superficial cells of the foot develop into haustorial protuberances. The initiation of an apical meristem in the shoot-region was not traced, although, judging from one particular embryo found, a single apical cell had apparently been set apart very early The position of the second apex of growth in the young plantlet was described. I have since been able to study a much larger number of embryos, although the series is still not quite complete, lacking certain stages as seen in transverse section. The present study confirms my previous conclusions, and makes more clear the main segmentations of the embryo. It also determines the early initiation of the apical meristem in the shoot, as well as that of the latter's secondary apex of growth.

The majority of the large number of prothalli which I have examined in external appearance apparently bore no embryos at all. On the other hand, a few, and they almost always of the stouter type, showed the presence of several (Plate LXIII, fig. 1). A good number of very young embryos were found showing only the first one or two segmentations, some of the stouter prothalli bearing from two to five of these. In most instances a prothallus did not bear more than one developing embryo, although one or more undeveloped fertilized archegonia might be present. This condition of things may be compared with what I have found in the prothalli of those New Zealand species of Lycopodium which belong to the two large subterranean types. For example, one large prothallus of L. fastigiatum, which conforms to the clavatum type, bore no fewer than eleven young embryos as well as three young plants. This was, of course, an exceptionally large number, but many of the prothalli of L. volubile, L. fastigiatum, and L. scariosum which I have sectioned showed three or four developing embryos, and it is quite usual for these prothalli to be found with two or three well-grown young plantlets attached to them.

Generally speaking, all stages in embryo development except the very youngest can be detected in external examination. The two general features which make them thus apparent are, firstly, a superficial localized, swelling of the prothallial tissue, and, secondly, the presence of an old archegonium neck at the apex of this swelling. These were what one always looked for. In the case of fairly well advanced embryos, which, however, had not as yet ruptured the prothallial tissues, the interior of this swollen region always appeared somewhat darker than the surroundings tissues.

It may be as well to state at once the most prominent features in the embryo of Tmesipteris. These are, firstly, the basal wall, which can clearly be traced throughout the whole development until the young sporophyte becomes detached (the plantlet usually detaching itself here, leaving its foot embedded in the prothallial tissues); secondly, the superficial swelling of the prothallial tissues around the embryo, together with the repeated transverse divisions of the large prothallus-cells lying immediately interior to the

– 395 –

foot (the browned “cup” being an outstanding feature on the prothallus-surface at an old point of attachment of a plantlet); thirdly, the haustorial protuberances from the foot into the tissues of the prothallus; fourthly, the presence of a single large apical cell in the outer, or shoot, region, in some cases a second apical cell being set apart in the other outer quadrant; and, lastly, the Tmesipteris embryo, when compared with those of other Pteridophytes, shows the important feature of the absence of suspensor, cotyledon, and root organs.

The hypobasal region of the very young embryo curves somewhat as it develops, so that frequently a single longitudinal section does not show both foot and shoot cut truly medianly. On this account, in illustrating some of the embryos on which my description is based, I have thought it advisable to show a series of several consecutive sections. Unless otherwise indicated, a series so illustrated always consists of consecutive sections. Again, in the epibasal region the apical cell never seems to be in the line of the archegonium-axis, so that the young shoot-apex bursts out from the tissues of the prothallus inclined at a greater or less angle, which, moreover, is not infrequently out of the plane in which the prothallus-limb lies. Hence in longitudinal section the apical cell is sometimes cut slightly obliquely.

In several cases developing embryos were seen on sectioning to be browned, the nuclei being small and the cell-walls more or less distorted. This would seem to have been due not to anything in the preparation of the material for embedding in paraffin, but to the previous death of the embryo.

B. First Segmentations.

In my earlier paper I described and figured several very young embryos (7, figs. 52 to 57), noting (p. 22) that the first wall to be formed divides the embryo into lower and upper regions, and that the next division is in the hypobasal cell by a wall leading at an angle from the basal wall into the lower end. The exact sequence of the subsequent segmentations was not demonstrated, although from the figures and from the further study of the same embryos it would appear that an inclined wall is formed also in the epibasal cell, the embryo thus attaining a quadrant stage. From the present study, also, this seems to be the normal sequence of segmentation, although several abnormal cases will be described.

Picture icon

Figs. 28a, 28b, 29.—Two fertilized egg-cells in longitudinal section. × 100.

The fertilized egg-cell at once grows considerably in size. A good many instances of this condition were observed, two of which are shown in longitudinal section in figs. 28a, 28b, and 29. The former of these was cut a little obliquely, but from fig. 28a it is apparent that cell-divisions in the surrounding prothallial tissue begin immediately. A considerable number of embryos showing the first segmentation only were found (figs. 30 to 34 a—c. The basal wall is always transverse, and divides the fertilized egg

– 396 –

into two more or less equal portions I have never found any variation from this. The surrounding prothallial tissue at this stage projects considerably (fig. 34), so that it is possible sometimes to detect these young stages in an external examination of the prothallus (see Plate LXIII, fig. 1). Following this, there is formed in the inner or hypobasal cell an inclined wall leading from the basal wall down towards the lower end of the embryo and dividing the hypobasal portion into two somewhat unequal quadrants. This stage is shown in longitudinal section in the series given in figs. 35a to 35d and figs. 36a to 36d, and in transverse section in the series figs. 37a to 37f. Next, a similarly inclined wall leads off from the basal wall towards the upper end of the epibasal cell, though not into

Picture icon

Figs. 30–34.—Five young embryos in longitudinal section, showing first division-wall only. The series 34a to 34c consists of consecutive sections, as in all series of sections illustrated in this paper unless otherwise stated. × 100.

the actual “beak,” the embryo thus attaining the complete quadrant stage. This is shown in longitudinal section in the two series figs. 38a and 38b, and figs. 39a to 39d, and in obliquely transverse section in the series figs. 40a to 40g. This sequence in segmentation seems to be the normal rule, so that before referring to the abnormal cases met with I will describe the subsequent cell-divisions which lead up to the setting apart of an apical cell in the epibasal region.

I have not found a sufficient number of young embryos cut transversely to determine whether or not there is normally a regular octant formation, but judging from the embryo cut obliquely transverse and illustrated in the series figs. 41a to 41g, and from others also, I should say not. In this case in the lower portion of the hypobasal region the first

– 397 –

inclined wall is alone present (figs. 41a to 41c). The section marked D probably also represents the embryo in section below the basal wall, showing a wall in each of the hypobasal quadrants which has led off from

Picture icon

Figs. 35a–35d, 36a–36d.—Two young embryos in longitudinal section, showing basal wall and also first hypobasal wall. × 100.
Figs. 37a–37f.—A young embryo in transverse section from below upwards, showing first hypobasal wall. × 100.
Figs. 38a, 38b, 39a–39d.—Two young embryos in longitudinal section, showing basal wall and also first epibasal and hypobasal walls. × 100.

the first inclined wall. In the epibasal region also (figs. 41e, 41f) the segmentation is not octant-wise. Figs. 42a and 42b show a slightly older-embryo in longitudinal section, in which a regular segmentation of the

– 398 –

quadrants in both epibasal and hypobasal regions has proceeded Those illustrated in figs. 54 to 57 of my previous paper correspond fairly closely with this. The first inclined walls in both main regions of the embryo shown longitudinally in figs. 43a to 43e are apparent, but the subsequent segmentation has followed a somewhat unusual course.

Picture icon

Figs. 40a–40g.—Young embryo in oblique transverse section from below upwards at a similar stage of development to those shown in figs. 38 and 39. × 100.
Figs. 41a–41g.—Young embryo in oblique transverse section, illustrating absence of octant walls. × 100.
Figs. 42a, 42b.—Young embryo in longitudinal section, showing first segmentations of the epibasal and hypobasal quadrants. × 100.

Abnormally segmented embryos are set forth in figs. 44a to 44c, 45a to 45f, and 46a to 46d. In each of these the basal wall is clearly to be distinguished, and also the first inclined wall in the hypobasal half, but in the two last-mentioned embryos the segmentation in the epibasal half is rather difficult to interpret. In that shown in figs. 44a to 44d a single

– 399 –
Picture icon

Figs. 43a–43e.—Young embryo in longitudinal section, showing irregular segmentation of epibasal and hypobasal quadrants. × 135.
Figs. 44a–44d.—Young embryo in longitudinal section, showing abnormal segmentation. In that marked D the transverse section of the prothallus is shown at the point where the embryo was borne. A to C × 180; D × 66.
Figs. 45a–45f.—Young embryo in longitudinal section, showing abnormal segmentation. × 135.

– 400 –

division of the epibasal half has taken place by a wall which, instead of being inclined to the basal wall, is parallel to it. The uppermost cell thus has almost the appearance of a suspensor, but comparison with other embryos shows that this cannot be its nature. Fig. 44d represents a transverse section of the parent prothallus at the point at which this embryo was borne, and it will be seen that the prothallial tissues protrude here rather more than usual. Possibly the embryo has been stimulated by this to a rapid elongation. In the other two abnormal cases mentioned the archegonial neck appears towards the end of each series, so that the sections must be considered more or less obliquely longitudinal. In the case of that shown in figs. 45a to 45f there seem to be two inclined walls leading off at very slight angles from the basal wall into the epibasal region, and along with this it must be noticed that the embryo is squat in form. It was situated well up the prothallial protuberance which surrounded the foot of a well-grown plant, where the cells, although not compressed, were

Picture icon

Figs. 46a–46d.—Young embryo in longitudinal section, showing abnormal segmentation. × 135.

yet all much extended in a horizontal direction. I would suggest that the extension of the young embryo in this direction had caused it to repeat the formation of the epibasal inclined wall. In the third of these abnormal cases (figs. 46a to 46d) it will be seen by reading the series from the last section backwards that the first-formed epibasal wall approaches the basal wall and presumably joins it before the section marked A is reached. Here too, then, it is apparent that this wall is inclined at an unusually slight angle, as is also that in the hypobasal region. These were the only abnormally segmented embryos observed.

C. Initiation of the Shoot Apex.

An apical cell is set apart comparatively early in one of the epibasal quadrants, and from this the shoot-apex is formed. In my earlier paper I noted that an apical cell was probably already present in the young embryo shown there in figs. 55 and 56, and a re-examination of these sections in the light of my subsequent studies confirms this belief. The

Picture icon

Fig. 1.—Tmesipteris. Photograph of a portion of a prothallus showing a protruding embryo and also two very young embryo protuberances at A and B. That at B contains two young embryos.
Fig. 2.—Tmesipteris. Photograph of the forward end of a prothallus, showing an embryo bursting through the prothalhal tissues.
Fig. 3.—Tmesipteris. Photograph of the forward end of a prothallus, showing a young attached plantlet with two apices of growth.

– 401 –

young embryo figured in longitudinal section in the present paper in the series figs. 47a to 47d possesses an apical cell from which segments have apparently already been cut off. In the series figs. 48a to 48f another embryo, at a slightly older stage of development, is given in longitudinal section which apparently possesses two apical cells, one in each epibasal quadrant. In these early stages it is not clear whether or not the cells alongside the apicals have actually been cut off from them as segments or have arisen simply by the general segmentation of the quadrants. In those figured the former seems to be the case, and the apicals are strongly defined Until the shoot has reached the size when the overlying prothallial tissues are ruptured, the apical cell usually cuts off more segments towards the base than towards the apex of the embryo. That illustrated

Picture icon

Figs. 47a–47d.—Embryo in longitudinal section, showing initiation of the apical cell and also the main divisions. × 100.

in fig. 48 is not cut medianly for foot and shoot together. The full size of the former appears at the beginning of the series, but of the latter in the sections marked D and E. The section F, which shows the neck of the archegonium, lies five sections beyond that marked E. In fig. 49 another embryo is shown in which the single apical cell has been functioning for only a short time. The position of this embryo in the transverse section of its parent prothallus is shown in fig. 50. A median longitudinal section of another such embryo, with the foot rather more developed, is given in fig. 51. In all these the basal wall and the first inclined walls in both epibasal and hypobasal regions can be distinguished. They illustrate also the beginning of the outgrowth of the superficial cells of the foot, and in fig. 49 the cell-divisions in the prothallial tissue abutting on the foot can be well seen.

– 402 –
Picture icon

Figs. 48a–48f.—Embryo in obliquely longitudinal section, showing two apical cells The section marked F is the fifth beyond that marked E. × 100.

– 403 –

Three obliquely transverse sections through the shoot region—the third, sixth, and ninth respectively from the one which first touches the top of the embryo—are shown in figs. 52a to 52c. The apical cell appears in

Picture icon

Fig. 49.—Embryo in longitudinal section, showing the apical cell and the main divisions. × 100.
Fig. 50.—Transverse section of the prothallus, giving the position of the embryo shown in fig. 49. A is the area in which rapid cell-division is taking place, and B is the fungal area. × 40.
Fig. 51.—Embryo in longitudinal section, showing apical cell and main divisions. × 100.
Figs. 52a–52c.—Three obliquely transverse sections from above downwards (Nos. 3, 6, and 9) through the epibasal region of an embryo. × 100.

that marked A. What is probably the first inclined division in the epibasal region appears in all three sections at aa, but an intersecting octant wall could not be traced throughout the series.

– 404 –

Two browned and probably dead embryos are illustrated in longitudinal section in figs. 53 and 54. In neither does the apical cell appear, but the main basal wall is obvious. The foot of that shown in fig. 54 had grown considerable.

Picture icon

Figs. 53, 54.—Two browned and dead embryos in longitudinal section, showing the main basal wall but not the apical cell. × 75.
Figs. 55a, 55b.—Embryo in longitudinal section, showing two apical cells in the shoot-region. A and B are not consecutive sections. × 75.
Figs. 56a, 56b.—Embryo in longitudinal section, showing two apical cells in the shoot-region. The sections A and B are not consecutive. × 75.

– 405 –

Not infrequently two apical cells, one in each epibasal quadrant, are set apart more or less simultaneously at an early stage. The youngest embryo which showed this feature is that in fig. 48, the apicals lying alongside one another at the apex of the shoot, separated only by the quadrant wall. Three other embryos, at rather older stages of development, which possess two apicals, are shown in figs. 55a and 55b, 56a and 56b, and 57a and 57b. While at first segments are cut off from the apicals rather towards the base of the embryo than outwards, all-round segmentation soon begins, and they become more widely separated, inclining from one another, as in the figures, at an obtuse angle, or even eventually in exactly opposite directions. In the Tmesipteris embryo the apical cells are always large and are readily observed, the regular arrangement of cells cut off from them being also a distinguishable feature. The growth of the young shoot from two similar apices will be dealt with in the next section of this paper, but the fact that the two apices are sometimes present together in the young embryo is noteworthy.

When only one shoot-apex is present a certain amount of cell-division takes place in the other quadrant until the young apex has actually burst through the prothallial tissue. The second quadrant thus forms a smooth rounded base to the shoot proper, consisting eventually of a uniform tissue of large-sized cells in which the symbiotic fungal coils early establish themselves. Before it emerges from the surface of the prothallus the shoot is more or less globular in shape, but the apex or apices soon become beak-like in form (Plate LXIII, fig. 2, and fig. 58a). A strand of elongated and narrow conducting-elements is early differentiated at the centre of the epibasal region by the longitudinal division of the cells there situated (figs. 56a, 56b, 57a, 57b). As the apex grows forward these narrow elements curve round and lead up behind it, extending back almost to the main basal wall When there are two apices present the two strands both lead down in this way towards the foot. The haustorial protuberances early arise all over the foot-surface by the outward growth of its superficial cells, and the foot as a whole sometimes assumes a very irregular shape (figs. 54, 56a, 56b, 57a, 57b). The full development of these outgrowths is not attained until the young plantlet has become well advanced. Starch is often present in the foot and central cells of the embryo in large quantities, and in the cells of the prothallus also which lie adjacent to the foot (figs. 56a, 56b). On account of the rapid cell-divisions, and also of the large size of the nuclei in the upper region of the young embryo, mitotic figures can often be seen here to great advantage.

The Young Sporophyte.

A. The Rhizome.

The forward growth of the young plantlet after it has emerged from the prothallial tissues is illustrated in figs. 58a, 59a, and 59b. In the former of these there is only one apex, and on account of its lower part not being cut medianly the conducting-strand does not appear. In the latter there are two equally-developed apices of growth, each with its conducting-strand. This plantlet had become detached from its prothallus. It may be compared with that shown in fig. 65 in my previous paper. Although for a considerable time the young plant is dependent upon its parent prothallus for the main food-supply, as evidenced by the continued extension of the haustorial outgrowths of the foot and the presence of

– 406 –
Picture icon

Figs. 57a, 57b.—Embryo in longitudinal section, showing two apical cells in the shoot-region. The sections A and B are not consecutive. × 100.

– 407 –

starch in and around them, it early forms rhizoids and shows the presence of the fungal coils in its cells. The largest of the embryos borne on the prothallus shown in Plate LXIII, fig. 1, may be compared with that in figs. 58a and 58b.

Young plants up to 4 mm. in length may frequently be found in which growth is taking place from only one apex. The base of the young stem is smooth and round, and in longitudinal section is seen to consist of

Picture icon

Figs. 58a, 58b.—A protruding embryo in longitudinal section, showing the beak-like apex, and also the presence of the endophytic fungus. The sections A and B are not consecutive. × 75.
Figs. 59a, 59b.—A very young detached prothallial plantlet in longitudinal section, showing two apices, each with its conducting-strand. The sections A and B are not consecutive. × 75.

a uniform tissue. There is no undeveloped apex present at this point. Two such plantlets are figured in my earlier paper (7, figs. 60, 67), and those shown at figs. 74 and 77 in the present paper will serve to illustrate the same point. In the majority of cases when the second apex of growth is formed in such plantlets it arises at the base of the first in just the position it would occupy if it had been initiated in the second epibasal quadrant of the young embryo. This second apex is inclined at a varying

– 408 –
Picture icon

Fig. 60.—The base of a young prothallial plantlet in longitudinal section, showing an early stage in the development of a secondary apex. × 75.
Fig. 61.—A young prothallial plantlet in longitudinal section, showing the secondary apex in an unusual position. × 45.
Figs. 62, 63.—Two young prothallial plantlets in longitudinal section, showing two equally-developed shoot-regions. × 55.

– 409 –

angle to the primary shoot, being sometimes almost in a straight line with it (figs. 73, 76, 78, 81, 83). Comparison may be made with those illustrated in my other paper in figs. 68, 69, and 71. I have not observed the actual initiation of this secondary apex when thus late developed, but fig. 60 represents an early stage. It is, of course, adventitious in origin, and, judging from what takes place in the case of the origin of lateral adventitious shoots on both old and young rhizomes, an apical cell is cut out from one of the surface cells while at the same time the inner cells lying between this and the vascular strand of the primary shoot divide longitudinally to form conducting-elements. I have observed a few instances out of the large number of plantlets examined in which the

Picture icon

Fig. 64.—The point of attachment of the young prothallial plantlet shown in Plate LXIII, fig. 3, in longitudinal section, showing foot, basal wall, and accumulation of starch. × 100.

secondary apex was not situated at the base of the primary stem, but much higher up. One such plantlet is shown in fig. 61 in longitudinal section. In this the appearance is rather as if there had been a dichotomy of the apex. However, I have never come across an undoubted instance of such dichotomy in a young rhizome, although it may be seen in older rhizomes. Plantlets in which two primary apices of growth are present are shown in longitudinal section in figs. 62, and 63, and in general view in Plate LXIII, fig. 3, and in figs. 75 and 83. A corresponding instance was given in my previous paper (fig. 70). Fig. 64 is a longitudinal section of the point of attachment of the young plant shown in Plate LXIII, fig. 3.

– 410 –
Picture icon

Figs. 65a–65c.—The young lateral adventitious apox shown in fig. 65d in three transverse sections from the apex downwards (Nos. 1, 2, and 5). × 75.
Fig. 65d.—Outline of a young piothallial plantlet in longitudinal section, showing the first and second apices of growth, foot, distribution of fungus, and the position of a lateral adventitious apex. × 20.
Figs. 66a, 66b.—The apical region of a young prothallial plantlet in longitudinal section, showing a very young lateral apex out obliquely. The sections A and B are not consecutive. × 75.
Figs. 67a–67c.—A very young lateral adventitious apex in three transverse sections from the apex downwards (Nos. 1, 7, and 12). × 75.

– 411 –

Adventitious branches are a well-known feature in older rhizomes of Tmesipteris, where they sometimes apparently function as storage-tubers before developing further. Laterally - developed adventitious apices may be found also in quite young plantlets (figs. 74, 75, 81). One such was present at the point indicated on the plantlet shown in longitudinal section in fig. 65d. It projected very slightly above the surface of the main stem, and is shown in transverse section in figs. 65a, 65b, and 65c, which represent the first, second, and fifth sections passing through it. A slight strand of narrow elements led from behind it to join the strand of the main shoot. Another young adventitious apex occurring in a similar position is shown in figs. 66a and 66b. In the section marked B the apex and strand of the main shoot is cut longitudinally, but the adventitious apex and its strand does not lie quite in the same plane. The apical cell of the latter appears in the seventh section from B, and is shown cut obliquely at A. In figs. 67a and 67b a very young lateral apex is shown cut transversely. The section marked A passes through the apical cell. This has evidently been functioning for some time, judging by the arrangement of the cells in B, which lies six sections below A. Deeper down towards the main strand, however, the adventitious strand has the appearance, as seen in C, as if it had arisen not from the apical cell, but by the subdivision of an ordinary cortical cell of the main shoot. Sometimes a plantlet will show a third apex of growth at its base in close proximity to the second apex, as illustrated in longitudinal section in fig. 68. Here the main strand has been cut obliquely transverse, since the foot into which it leads lies in a plane at right angles to the direction of growth of the two young apices. The latter also are not cut medianly throughout their length, so that the course of their strands is not included in the figure. A plantlet in a similar condition is also shown in general view at fig. 79. One very young plantlet (figs. 69a, 69b, 69c) was found on sectioning to have three apices. Two of these—namely, B and C—had given rise to well-defined strands, and had probably been initiated in the embryo. The third, shown at A, had given rise as yet to no strand, and lay rather out of the plane of the other two, as can be seen from the fact that this section does not include the foot. It must probably be interpreted as an adventitiously-formed apex rather than as one which had arisen in the embryo.

Picture icon

Fig. 68.—A young prothallial plantlet in longitudinal section, showing the primary apex and also two apices at the base of the plantlet. × 45.

The apical cell of the main shoot in the young subterranean plantlet, and its manner of segmentation, is shown in longitudinal section at fig. 71. A series of transverse sections taken at intervals from apex to foot through a young plant of about the same age as that shown at Plate LXIII, fig. 3,

– 412 –
Picture icon

Figs. 69a–69c.—A very young attached prothallial plantlet in longitudinal section, showing three apices of growth, two of which possessed conducting-strands. Sections B and C are respectively the tenth and fifteenth from A. × 47.
Figs. 70a–70f.—A young prothallial plantlet in six transverse sections taken at different points from the apex downwards to the foot. A to D × 100; E and F × 80.

– 413 –

is given in figs. 70a to 70f. The section marked A passes through the apical cell of the shoot, and shows that it segments regularly from four sides. B, C, and D show the differentiation of the strand in progressively older regions of the shoot. Section E is taken at a point a little above the main basal wall, and shows the narrow conducting-elements which lead down

Picture icon

Fig. 71.—The apex of a young prothallial plantlet in longitudinal section, showing the apical cell and its segmentation. × 75.
Fig. 72.—The point of attachment of a young plantlet to the prothallus, showing unusually long intraprothallial shoot-region. The extra-prothallial shoot-region is cut obliquely. × 80.

to it from the shoot. Section F shows the foot in transverse section a little below the basal wall, the original first inclined wall in the hypobasal region of the embryo being still very evident. The haustorial outgrowths from the foot of a well-grown plantlet are illustrated in my previous paper at Plates II and III and figs. 58 and 59, the main basal wall appearing in the latter figure. In fig. 72 in the present paper is shown a young plantlet

– 414 –
Picture icon

Figs. 73–87.—Young subterranean plantlets in general view, showing variations in form. pr. = prothallus; ft. = foot: adv. ap. = lateral adventitious apex.

– 415 –

in longitudinal section, in which that part of the epibasal region contained within the prothallial tissues was of unusual length. The extra-prothallial shoot-region is cut somewhat obliquely, so that the course of the strand becomes lost.

From the above account it will be seen that the young sporophyte of Tmesipteris, before the development of the aerial shoot, shows variations in form. A number are given in figs. 73 to 87, and with these can be compared others illustrated in my previous paper at figs. 66 to 72. The development of a third apex of growth has given an irregular form to those shown at figs. 80, 85, 86, and 87. In the plantlet at fig. 82 probably the longer of the two branches was the one secondarily developed, and it here occupies an unusual position. Some of these figures show that the plantlet may attain a considerable size while still attached to its prothallus. When detached they generally show a fragment of old prothallial tissue still attached to the foot, frequently in the form of a dark ring. The plantlet apparently becomes detached from the prothallus at the basal wall, and sections through a prothallus at an old point of attachment invariably show the whole foot of the plant still embedded in its tissues.

It may be stated here that throughout the life of the sporophyte no indications are to be met with of the adoption of any special root-like function on the part of any of the branches of the rhizome. These branches are all similar to one another in both external appearance and internal structure.

B. The Aerial Shoot.

The young wholly-subterranean plantlet frequently attains a length of ½ in. to ¾ in. before forming an aerial branch. Generally one of the two main growing apices turns up out of the soil, the other continuing to extend in the humus (figs. 88, 89). In some cases both ends may grow out into aerial shoots (fig. 90, and 7, fig. 73), the rhizome-system then extending by the formation of lateral branches. Again, in other instances, the first aerial branch arises laterally, the main apices of the rhizome continuing underground (7, fig. 5).

The aerial shoot is much thinner than the rhizome, and is at first quite scaleless and leafless. Usually when the shoot is from ½ in. to ½ in. high, leafy outgrowths are formed immediately behind the apex, but these form only scale leaves. The first aerial shoot generally does not grow more than 1 in. or 2 in. in height, and remains very slender and sterile, withering off when other shoots are formed. Frequently the second may do the same; but those next formed are much longer, although still slender. The mature well-grown shoots are to be found only when the rhizome-system has become strongly developed. In most cases aerial shoots remain unbranched, but a single forking sometimes takes place at or near the base, or occasionally, in well-grown pendulous branches, even higher up.

At the base of well-grown aerial shoots there is generally only a short region bearing the scattered scale leaves, the ordinary form of the leaf being fairly early and often suddenly attained, but the first shoots of the young plant frequently show a much longer scale-leaf region. In the latter the transition to the larger form may be either sudden or gradual. There is much variation in the size shown by the mature sterile leaf, this generally being longer, as migght be expected, in pendulous branches than in those of more erect growth. The leaves of juvenile plants, also, are rather small in size. Sometimes however, elongated pendulous branches show a much

– 416 –

shorter form of leaf than usual. Three examples of mature sterile leaves are given in figs. 91a to 91c, the first two coming from luxuriantly-growing pendulous plants, and the third from the lower part of a short, erect, but fully-grown shoot. The normal size is narrower and slightly longer than that shown in 91c. Individual branches may be met with showing marked

Picture icon

Figs. 88, 89.—Young plantlets, showing first aerial shoot. × 3.
Fig. 90.—Young plantlet in which both main apices of rhizome have grown up into aerial shoots. × 3.
Fig. 91.—Three varieties of the mature sterile leaf taken from well-grown shoots. × 1.3.

variation in size of the leaves up and down the branch, a zone of quite stunted and almost scale-like leaves sometimes occurring in amongst those of the ordinary form. The leaves are sometimes in two orthostichies only, being then flattened in one plane, this being a common state in juvenile

– 417 –

slender branches, but to be seen also in older ones. Again, they may be arranged in three or in four orthostichies, the stem in transverse section showing a corresponding number of ridges. These different leaf-arrangements may be found intermixed along the one branch. The leaves are never in whorls, but are scattered.

In pendulous branches the sporophylls are grouped in zones alternating with sterile zones. A characteristic very short and compact variety may sometimes be found on tree - ferns in which the whole of the upper two-thirds of the branch is fertile, there being no zoning. The two or three first-formed small aerial shoots in the young plant remain sterile. Sporophylls generally make their first appearance singly, juvenile shoots about 4 in. to 6 in. in height commonly showing one sporophyll situated about half-way along their length, and sometimes also another

Picture icon

Fig. 92—Portion of sporophyll, showing abnormal number of loculi in the synangium. × 1.5.
Figs. 93a, 93b.—A sterile leaf in side and face views, showing a lobe-like outgrowth from the lamina. × 3.
Fig. 94.—The upper portion of a sterile leaf showing a small marginal outgrowth and thickening. × 3.
Figs. 95a, 95b.—The apex of a sterile leaf in side and face views, showing forking of the tip. × 3.

borne singly still nearer the apex. Well-grown shoots showing the usual alternate fertile and sterile zones may be found with these single sporophylls towards the base. Occasionally a juvenile shoot is found in which the sporophyll formation has been initiated not singly but in a normal zone. The lobes of the sporophyll are similar in form, and sometimes also almost in size, to the sterile leaf appendages, but generally they are narrower.

Abnormalities in the sporophylls have been described by A. p. W. Thomas (15). I have occasionally found synangia with three loculi instead of two (fig. 92), and not infrequently it happens that one of the lobes of the sporophyll is more or less reduced in size, or is even almost entirely absent. A sterile leaf in one instance bore about half-way down its flat-surface a small lobular outgrowth (figs. 93a and 93b), the appearance being that of an abortive forking. Both sterile leaves and the lobes of the sporophyll sometimes show a slight projection at the margin with the tissues of the lamina thickened immediately behind it (fig. 94), suggesting a less-developed stage

– 418 –

of the outgrowth shown in fig. 93. At the tip of another sterile leaf an outgrowth was present alongside the mucro, as if the tip was preparing to fork (figs. 95a and 95b). Such abnormalities may or may not have any significance in indicating that reduction has taken place in these leafy appendages, but they seem to show that the sporophyll and the sterile leaf are not in nature essentially different from one another, and that neither of them is altogether fixed in form. In fact, the sporophyte generation as a whole in Tmesipteris provides many details indicating, as might be expected in a plant showing undoubted primitive characters, a general lack of specialization.

General Considerations.

As has been noted above, there is a very striking similarity in form and general external appearance between the prothallus and the young subterranean sporophyte of Tmesipteris. It is evident that this similarity between the two generations holds for Psilotum also, judging from Lawson's description (12) of the prothallus and that of Solms Laubach (14) with respect to plantlets developed from buds on old rhizomes. Bearing in mind how plastic, generally speaking, the Pteridophyte gametophyte is known to be, and, moreover, how largely the distribution and persistence of these ancient families has probably been due to the ability of the gametophyte to adapt itself to new conditions, it would be, of course, unwise to conclude too hastily that the similarity between the two generations in the Psilotaceae is a primitive feature. On the other hand, a close correspondence of this nature is not found in the life-history of other modern Pteridophytes, even in those epiphytic forms of Lycopodium and Ophioglossum which possess a cylindrically-built, branched, and ramifying prothallus; nor can this be attributed altogether to the more complex organization of the young sporophyte in these families. Whatever view we take of this external similarity as it is to be seen in the Psilotaceae, the fact that it exists is at least worthy of attention, and becomes even more so when it is found to go hand in hand with certain important structural features in both generations which can with reason be claimed to be primitive.

The superficial position of the sexual organs on the prothallus of Tmesipteris and Psilotum can be regarded as a structural feature of the gametophyte which has not arisen by modification from a more deeply situated position. The adoption of the subterranean habit of growth in other pteridophytic families has not resulted in a similar simple organization and structure of the sexual organs as is to be found in the Psilotaceae, and it is therefore difficult to see why in the latter this simplicity should be regarded as the result of modification. The persistence of the single apical cell throughout the life of the prothallus, the dichotomous branching, the gradual extension in girth of the prothallus from an initial filament without the formation of such a primary tubercle as is found in some Lycopodiums, and the complete absence of any differentiation of tissues in the prothallus-body, may all be urged as more or less primitive features.

In correspondence with the superficial position of the sexual organs on the prothallus, the embryo also is shallowly seated, there being no suspensor organ to push it down into the food-supplying tissues of the prothallus such as has been developed in the Lycopodiaceae, and in certain also of the Ophioglossaceae. On the assumption that the suspensor is a primitive organ, it might be urged that it had become lost in Tmesipteris owing to

– 419 –

the very early adoption of the fungal habit by the young sporophyte and its consequent ability to nourish itself. However, the dependence of the young sporophyte upon its prothallus is a protracted one, and the absence of the suspensor is compensated for by the development of haustorial protuberances from the foot, just as is found, only there not to so great an extent, in the sporogonium of the Anthoceroteae. The inference seems to be that the superficial position of the embryo and the absence of the suspensor is the more primitive condition.

There are only two main body-organs in the embryo and young sporophyte of Tmesipteris—namely, the shoot and the foot—there being no trace of root, cotyledon, or suspensor. Thus its embryogeny is the simplest among existing Pteridophytes. It would be difficult to conceive of a more simple organization for a vascular cryptogam, and in instituting comparisons we are forced to look to the young sporogonium of Anthoceros rather than to any Pteridophyte embryo. While not suggesting that Tmesipteris has been actually derived from the Anthoceros cycle of affinity, it is clear that the absence from the former of any such organs as root or cotyledon suggests that they approximate in so far as they both represent primitive lines of development. In his Origin of a Land Flora Bower contemplates the fundamental structure-plan of the various pteridophytic types of embryo as a spindle-shaped axis with the shoot-apex situated at the apex of the epibasal region. The embryo is primarily a shoot, and the other main body-organs are appendages developed secondarily upon it. Speaking of the light which it was hoped the embryogeny of the Psilotaceae would throw upon this matter, he says (ibid., p. 421), “If the embryo develops without appendages directly into the rootless and leafless rhizome, then either reduction has been effective back to the earliest phases of the individual, or the sporophyte at first represents that primitive state of an axis without appendages which a strobiloid theory contemplates in the far-back ancestry.” That the simplicity of Tmesipteris is not due to reduction is a belief which has been greatly strengthened by the discovery of the rootless and leafless Rhyniaceae. The embryogeny of Tmesipteris as described in the present paper makes more clear - cut the theory of the origin of the sporophyte of the Pteridophyta from an Anthoceros-like sporogonium.

Pursuing this theory further, it may be noted that two definite suggestions based on the embryogeny of existing Pteridophytes have been put forward as to how this origin could have taken place. Campbell (2, p. 210) would see in the young embryo of Ophioglossum moluccanum a primitive type of Ophioglossum which, can be derived from an Anthoceros-like ancestor. In this species the lower portion of the embryo forms the large foot and the upper the cotyledon, the latter, however, not being sporogenous, as is the upper part of the sporogonium of Anthoceros. The new organ in Ophioglossum is the root which arises at the junction of the cotyledon and the foot. There is at first no stem-axis, this being developed late as a secondary structure upon the primary root. Campbell links up Ophioglossum, which he regards as “the most primitive type of the fern series” (ibid., p. 42), with the Bryophytes in the suggestion that the Anthoceroteae progressed to the formation of a root from the basal meristem of the sporogonium, and that the “pro-Ophioglossum” produced spores upon the first leaf. Bower (1, p. 469) has criticized this theory by pointing out that for his primitive form Campbell has chosen the most abnormal of all the species of Ophioglossum, instead of starting with such

– 420 –

a form, as O. vulgatum, which approximates more to the usual pteridophytic type of embryo.

The presence of the “protocorm” in the embryo plant of certain species of Lycopodium has given rise to another suggestion with regard to the origin of the free-living sporophyte—namely, that the protocorm was the precursor of the leafy shoot. Against this it has been urged, especially by Bower (1), that the embryo of Lycopodium is prone to parenchymatous swellings, and that the protocorm is best regarded, as a physiological specialization. My own study of the large development of this organ in the two New Zealand species L. laterale and L. ramulosum (4, 5) led me to conclude that its abnormal size in these two species could be put in connection with the fact that the young sporelings were required to tide over a summer season, during which their natural boggy habitat would be usually dried up, before they could establish themselves, and that the manner of development of their protocormous rhizome was capable of a physiological explanation. I concluded that this lent weight to the theory that the Lycopodium protocorm in general may best be interpreted in this way (4, p. 289). I was careful, however, to add that the fact that this organ is characteristic of two out of the five sections of the genus Lycopodium, and is also present in a specialized form in Phylloglossum, indicates “a considerable degree of antiquity for the protocorm within the genus Lycopodium.” Vegetatively-produced plantlets of these New Zealand species possess a basal protocorm (5), as Osborn also (13) has shown in plantlets of Phylloglossum produced on detached leaves. This author inclines to regard the tuber of Phylloglossum as of physiological importance only. Since writing my first accounts of the Lycopodium protocorm I have found at the close of a dry summer season a colony of young sporelings of L. cernuum growing upon a roadside clay cutting in which this organ was as largely developed through the formation of a rhizomatous extension as in the other two New Zealand species mentioned (6, p. 189). This is unusual, for in L. cernuum the stemaxis is generally initiated early on the protocorm, and it indicates that the unusual conditions were the cause of this extra development. Recently Kidston and Lang (9) have described under the name Hornea Lignieri a rootless protocormous plant from the Early Devonian of Scotland which “retains in the adult condition an organization comparable to the protocorm stage in the species of Lycopodium. The relation of the aerial stems of Hornea, to the rhizome is similar to that of the protophylls to the protocorm in Lycopodium” (9, p. 620). It is, of course, a perfectly legitimate criticism to make that even in this archaic plant the protocorm is merely a physiological specialization called forth by precisely the same conditions as govern the life of the modern swamp-growing species of Lycopodium, and the fact that the authors have indicated (9, note, p. 612) that an intercellular fungus is present in the rhizome is of considerable significance in this respect. However, the plant in its general organization is obviously primitive, and is associated with other types of plants of very simple structure, and, belonging as it does to a group which comprises the earliest known land-plants, can be considered as lending great weight to Treub's theory of the protocorm. If the protocorm can be regarded as primitive, it is, of course, open to be interpreted either as an organ by which the supposed sporogonium-like ancestor of the Liycopods first attained independence of the gametophyte, or as a more or less modified representative of a possible thalloid ancestor.

– 421 –

A third definite suggestion with regard to the origin of the leafy axis of existing Pteridophytes from a strobiloid ancestor arises out of the facts of the simply organized embryo of Tmesipteris. The only new feature to be postulated here is the extension in length of the shoot from an apical meristem instead of, as in Anthoceros, from an indefinite basal meristem, and the initial cause of the continued shoot-elongation might be set down as being the adoption of a subterranean mode of life by the gametophyte. The differences between the ancestral strobilus and the derived rootless shoot would then be referred largely to their different modes of life. Further development in complexity of the sporophyte of this “pro-Tmesipteris” would take place by the continued growth of the shoot and by its dichotomous and lateral branching. On this view, the subterranean habit of the gametophyte would be regarded as of very early origin in at least one line of descent of the higher plants, although in other phyla, as, for example, that of Lycopodium, and possibly also that of the Ophioglossaceae, it is probably a very much later development. Bower in a general way regards the subterranean habit of the gametophyte as being modified from the subaerial habit. He says (1, p. 710), “It may accordingly be concluded as probable that the prothallus of early Pteridophytes at large was a relatively massive green structure with deeply-sunk sexual organs.” If, as suggested above, the superficial position of the sexual organs in the Psilotaceae is not a modified feature, the gametophyte of this class stands apart from that of other Pteridophytes. From Kidston and Lang's description of the asexual generation in the Rhyniaceae one is tempted to conclude that the gametophyte of these plants was subterranean rather than subaerial. The presence in the gametophyte of an endophytic fungus is a widespread feature of existing Pteridophyte prothalli, and I suggest that it may with as much reason be considered to have played a part in leading to the development of the rootless and leafless shoot of the Psilotaceae as to have been the cause of reduction taking place in a more complex plant-body. Modern Pteridophytes are so far removed in point of time from the hypothetical primitive form or forms that deductions based on comparative embryology, lacking as they do any support which might have been afforded by a knowledge of the embryogeny of archaic vascular plants, might be considered as altogether undependable. On the other hand, the extreme simplicity of the Tmesipteris embryo, wholly devoid as this is of appendicular organs, is full of significance, and the demonstration of a rootless and leafless condition in the earliest known land-plants strengthens the belief that the Psilotaceae have preserved in the first stages of their development primitive features.

The lateral origin of branches in the young rhizome of Tmesipteris would seem to be a more specialized character than branching by dichotomy of the apex, and it is curious to find that the latter does not apparently take place in the youngest rhizomes, whereas in those of a somewhat older age it is present along with lateral branching. The initiation of the second apex of growth may take place in the young embryo or be postponed till the shoot is well advanced, and even then varies in its position. This is just such a generalized character as might be expected in a primitive type of plant-body. The distinction also between the subterranean and the subaerial parts of the sporophyte would seem to be very indefinite, one or both of the first apices of growth emerging and becoming leafy according as the needs of the young plant direct. Sometimes the first-formed aerial shoot

– 422 –

is still more unspecialized in its origin, arising as a lateral branch from the rhizome just as do the aerial shoots in the older state. The alternation of fertile with sterile zones on the aerial branch is not a fixed character, a branch being zoned or practically altogether fertile according to its habit of growth. Neither the sporophyll nor the sterile leaf is fixed in form. Thus the Tmesipteris sporophyte is a peculiarly unspecialized plant-body not only in the absence of cotyledon and root organs from the embryo, but also in the general organization of the rhizome and of the aerial shoot.

In their description of the rootless and leafless Rhyniaceae, Kidston and Lang express the opinion that this simple plant-body “might as well be termed a cylindrical branched vascular thallus as a stem” (9, p. 619). They are more inclined to interpret it in the light of the theory that the sporophyte of the higher plants has arisen by modification and by specialization in the time of appearance of the asexual stage of an algal ancestor, rather than as the result of the adoption of an independent existence by a sporogonium-like ancestor with the consequence of a progression in sterilization of its parts. The simple organization of the Psilotaceae is, of course, susceptible of the same interpretation—but this has not been the one followed in the above remarks.

Literature cited.

1. Bower, F. O., The Origin of a Land Flora. Macmillan and Co., London, 1908.

2. Campbell, D. H., The Eusporangiatae: the Comparative Morphology of the Ophioglossaceae and Marattiaceae. Carnegie Institution, Washington, 1911.

3. Darnell-Smith, G. P., The Gametophyte of Psilotum, Trans. Roy. Sac. Edin., vol. 52, pt. i, pp. 79–91, 1917.

4. Holloway, J. E., Studies in the New Zealand Species of the Genus Lycopodium, Part I, Trans. N.Z. Inst., vol. 48, pp. 255–303, 1916.

5. —–, Ibid., Part II, Methods of Vegetative Propagation, Trans. N.Z. Inst., vol. 49, pp. 80–93, 1917.

6. —–, Ibid., Part III, The Plasticity of the Species, Trans. N.Z. Inst., vol. 51, pp. 161–216, 1919.

7. —–, The Prothallus and Young Plant of Tmesipteris, Trans. N.Z. Inst., vol. 50, pp. 1–44, 1918.

8. Kidston, R., and Lang, W. H., On Old Red Sandstone Plants, showing Structure, from the Rhynie Chert Bed, Aberdeenshire: Part I. Rhynia Gwynne-Vaughani, Trans. Roy. Soc. Edin., vol. 51, pt. ni, pp. 761–83, 1917.

9. —–, Ibid., Part II, Additional Notes on Rhynna Gwynne-Vaughani, with Descriptions of Rhynia major and Hornea Lignieri, Trans. Roy. Soc. Edin., vol. 52, pt. iii, pp. 603–27, 1920.

10. —–, Ibid., Part III, Asteroxylon Mackiei, Trans. Roy. Soc. Edin., vol. 52, pt. iii, pp. 643–80, 1920.

11. Lawson, A. A., The Prothallus of Tmesipteris tannensis, Trans. Roy. Soc. Edin., vol. 51, pt. iii, pp. 785–94, 1917.

12. —–, The Gametophyte Generation of the Psilotaceae, Trans. Roy. Soc. Edin., vol. 52, pt. i, pp. 93–113, 1917.

13. Osborn, T. G. B., Some Observations on the Tuber of Phylloglossum, Ann. Bot., vol. 33, pp. 485–516, 1919.

14. Solms Laubach, Der Aufbau des Stockes von Psilotum triquetrum und dessen Entwickelung aus der Brutknospen, Ann. du jard. bot. de Butt, vol. 4, pp. 139–94, 1884.

15. Thomas, A. p. W., The Affinity of Tmesipteris with the Sphenophyllales, Proc. Roy. Soc., vol. 69, pp. 343–50, 1902.